Identifying Quantum Structure in AI Language: Evidence for Evolutionary Convergence of Human and Artificial Cognition
A new research paper suggests that large language models like ChatGPT and Gemini exhibit quantum-like structures in their language processing, mirroring human cognition. The study found violations of Bell's inequalities and the presence of Bose-Einstein statistics in word distributions, indicating a non-classical probability model. Researchers propose this suggests an evolutionary convergence between human and artificial intelligence, with LLMs rapidly developing similar organizational principles for meaning in vector spaces. AI
IMPACT Suggests a fundamental similarity in how human and artificial minds organize meaning, potentially impacting future AI architecture and cognitive science.